US5636368A  Method for programming complex PLD having more than one function block type  Google Patents
Method for programming complex PLD having more than one function block type Download PDFInfo
 Publication number
 US5636368A US5636368A US08370768 US37076894A US5636368A US 5636368 A US5636368 A US 5636368A US 08370768 US08370768 US 08370768 US 37076894 A US37076894 A US 37076894A US 5636368 A US5636368 A US 5636368A
 Authority
 US
 Grant status
 Grant
 Patent type
 Prior art keywords
 function
 logic
 step
 subroutine
 equations
 Prior art date
 Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
 Expired  Lifetime
Links
Images
Classifications

 G—PHYSICS
 G06—COMPUTING; CALCULATING; COUNTING
 G06F—ELECTRICAL DIGITAL DATA PROCESSING
 G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
 G06F17/50—Computeraided design
 G06F17/5045—Circuit design
 G06F17/5054—Circuit design for userprogrammable logic devices, e.g. field programmable gate arrays [FPGA]
Abstract
Description
1. Field of the Invention
The present invention relates to programmable logic devices, and more specifically to a method for programming complex programmable logic devices having multiple function block types.
2. Background Art
Programmable logic devices (PLDs) are a class of integrated circuits (ICs) which can be programmed by a user to emulate various logic functions. Logic designers typically use PLDs to implement control logic in electronic systems because they are relatively easy to program, and often can be reprogrammed to change the emulated logic function. This makes their use in an electronic system's design phase less costly than custom hardwired or "application specific" integrated circuits (ASICs).
One major class of PLDs includes a set of input pins, a programmable AND plane connected to the input pins, an OR plane connected to output lines of the AND plane and a set of output pins connected to output lines of the OR plane. The AND plane provides a matrix of programmable connections where each column connects to an input pin and each row forms an output line of the AND plane, called a product term line, which connects to the OR plane. The OR plane may be programmable, such that each product term line is connectable to columns leading to different output pins, in which case the PLD is called a programmable logic array (PLA). Alternatively, the OR plane may be fixed, such that each product term line is assigned to a particular output pin, in which case the PLD is called a programmable array logic (PAL) device.
Because PLAs and PALs contain two levels of logic (AND and OR), these arrays are capable of implementing logic functions that are representable in a "sum of products" form. A sum of products form of a logic function is essentially a set of product terms (pterms) for each output of the function. Such a logic function is represented in a PLD by programmed connections in the AND plane and OR plane. Each pterm line has a programmable input connection in the AND plane to each input pin and produces a single output value representing the logical AND or "product" of the connected inputs. Usually, both the original input pin value and its complement are available for connection to a pterm line. Each output has a programmable pterm connection in the OR plane and produces an output value representing the logical OR or "sum" of the connected pterms.
These early PLDs were wellreceived by logic designers. However, as logic functions grew increasingly larger and more complex, logic designers were required to wire together two or more small PLDs to provide sufficient logic capacity. Although this process was tolerated during development and testing, it increased the cost and size of production units. This generated a demand for PLDs with increasingly larger logic capacity.
To meet the everincreasing demand for greater capacity, PLDs with increasingly complex architectures have been developed. One popular complex PLD type, known as erasable programmable logic devices (EPLDs), includes a plurality of "function blocks," input/output (I/O) resources, and an interconnect matrix such that each of the function blocks is interconnectable to any other function block or I/O resource of an EPLD through the interconnect matrix. Each function block of an EPLD is structured like the twolevel PLDs, described above. In effect, EPLDs incorporate several early PLDs and associated connection circuitry onto a single integrated circuit. This provides a circuit designer the convenience of implementing into a single device a complex logic function intended to program a wired collection of smaller PLDs using a single IC.
The function blocks of early EPLDs were essentially identical in size (capacity) and signal delay. Therefore, programming methods developed for these early EPLDs were primarily focused on implementing the desired logic functions into as few function blocks as possible, thereby leaving available as many resources of the EPLD as possible. One such prior art programing method is described in "A Fast Partitioning Method for PLABased FPGAs", IEEE Design & Test Of Computers, 1992, by Z. Hasan, D. Harrison, and M. Ciesielski.
More recently, a "Dual Block" EPLD was developed which includes two types of function blocksa "fast" function block (FFB) whose architecture, in effect, sacrifices programmability for speed, and a "high density" function block (HDFB) whose architecture is designed for maximum programmability. This type of EPLD allows portions of a logic function requiring fast inputpintooutputpin times to be implemented by the FFBs, and the remaining portions of the logic function to be implemented by the HDFBs.
The abovementioned prior art programing methods are inadequate for programing "Dual Block" EPLDs because these methods fail to consider the benefits of maximizing the use of one type of function block over the other.
The present invention involves programing complex PLDs having multiple function block types to implement logic functions, whereby the logic function are mapped into one of the function block types before being mapped into the remaining function block types. This allows desirable features of one function block type to be utilized by the logic function before other function block types.
In the specific example described below, an EPLD contains both "fast" function blocks (FFBs), which have limited programmability but exhibit highspeed signal propagation, and "high density" function blocks (HDFBs), which have a high degree of programmability but exhibit relatively slow signal propagation.
A programming method in accordance with the present invention initially converts the logic equations of a user's logic function into a form which has the greatest possibility of being mapped into either the FFBs or the HDFBs. Then, the equations are designated by the user as being "fast" (i.e., designated for implementation in an FFB). These designated equations are then divided (split) into subequations (if necessary) which can be mapped easily into the FFBs. The remaining (undesignated) equations are tested to determine if they can be implemented in the FFBs and, if so, they are split (if necessary) into subequations which can be easily mapped into the FFBs. However, if the undesignated equations cannot fit within the FFBs, they are split (if necessary) into subequations which can be easily mapped into the HDFBs.
The programming method then determines whether any of the equations of the logic function are implementable in the interconnect matrix linking the FFBs and HDFBs. A full crosspoint interconnect matrix acts as an AND gate. The programming method searches for equations which are implementable in as interconnect matrix AND functions, and tests whether the implementation of the equation in the interconnect matrix is possible and desirable. More particularly, the program tests whether the signals from the different function block types are of consistent polarities (positive or negative) to allow implementation in the interconnect matrix. Further, if implementation of the equation in the interconnect matrix causes a predetermined number of equations to be unimplementable in an FFB, then interconnect matrix implementation is not used.
The programming method then maps the converted and split equations of the logic function into the FFBs and HDFBs. In accordance with the present invention, the mapping process initially maps only the desired function block type (e.g. FFBs) until the desired function block type is fully utilized. In one embodiment, the program attempts to map every FFBqualified equation into each of the FFBs to assure that each FFB is fully utilized. Only after all of the FFBs have been maximally utilized or until all FFBqualified equations have been mapped does the program proceed to map the HDFBs. This process can be extended to secondary function block types in applications where there are three or more function block typesthat is, following the mapping of a mostdesirable function block type, the program proceeds with mapping a second mostdesirable function block type.
After mapping, the programming method performs a refinement subroutine with rearranges the mapping placement of the equations to produce an efficient arrangement of the logic function. The program recognizes that constraints are placed on the rearranging process based on the designation of FFB equations and the desirability of fully implementing the FFBs.
Finally, after the logic function is fully mapped, a bit map is formed which is downloaded by the computer to a device programmer, which applies the bit map to the individual programmable elements of a target EPLD, thereby programming the EPLD to implement the logic function.
These and other features, aspects and advantages of the present invention will become better understood with regard to the following description, appended claims, and accompanying drawings, where:
FIG. 1 is a circuit diagram showing a EPLD including multiple function block types;
FIG. 2(A) is a simplified schematic diagram of a Fast Function Block (FFB) of the XC7300 EPLD family;
FIG. 2(B) is a simplified schematic diagram illustrating a product term assignment associated with the Fast Function Block of FIG. 2(A);
FIG. 3 is a simplified schematic diagram showing a HighDensity Function Block (HDFB) of the XC7300 EPLD family;
FIG. 4 is a simplified schematic diagram showing an input/output structure of the XC7300 EPLD family;
FIG. 5 is a diagram showing a typical "machine" for programming an EPLD in accordance with the present invention;
FIG. 6 is a flowchart showing the basic subroutines in an EPLD programming method according to the present invention;
FIG. 7 is a flowchart showing a logic optimization procedure of the EPLD programming method;
FIG. 8 is a flowchart showing a logic splitting procedure of the EPLD programming method;
FIGS. 9A and 9B are flowcharts showing a UIM ANDing procedure of the EPLD programming method;
FIG. 10 is a flowchart showing a FFB Mapping procedure of the EPLD programming method;
FIG. 11 is a flowchart showing a HDFB Mapping procedure of the EPLD programming method; and
FIG. 12(A) and 12(B) are flowchart showing a Refinement procedure of the EPLD programming method.
The present programming method is directed to erasable programmable logic devices (EPLDs) incorporating two or more function block types. A "function block type" is generally defined as a family of substantially identical function blocks including a particular arrangement of logic circuitry and programmable connections. For example, function block types may vary by size (e.g., the number of input signals or generated product terms), by speed or level of programmability (e.g., the number and location of programmable elements used to configure signal paths within the function block), and by specialized circuitry (e.g., one function block type may include an arithmetic logic unit not found in other function block types of the EPLD).
The present programming method is described with particular reference to the XC7300 EPLD family manufactured by Xilinx, Inc. of San Jose, Calif. A brief description of the XC7300 EPLD family is provided below. Additional description of the XC7300 EPLD family is provided in Programmable Logic Data Book, 1994, published by Xilinx, Inc., which incorporated herein by reference in its entirety. However, note that the present invention is not limited to the specific function block types found in the devices of the XC7300 EPLD family.
Dual Block EPLD
FIG. 1 shows a simplified block diagram of an EPLD 100 which includes features common to the XC7300 EPLD family. The XC7300 EPLD family employs a DualBlock architecture including Fast Function Blocks (FFBs) 200 and HighDensity Function Blocks (HDFBs) 300. A Universal Interconnect Matrix (UIM) 400 provides programmable interconnections between the FFBs 200 and the HDFBs 300. Input signals of EPLD 100 are entered into HDFBs 300 and UIM 400 through Input/Output (I/O) pins 500A. Inputs signals are also entered into the FFBs 200 and UIM 400 through dedicated "fast" input pins 500B. Output signals are routed to I/O pins 500A from HDFBs 300, and to dedicated "fast" output pins 500C from FFBs 200.
Fast Function Blocks (FFBs)
FIG. 2(A) shows a simplified schematic of an FFB 200. Each FFB 200 includes an AND array 210 and nine macrocells (one macrocell 220 shown). The AND array 210 receives 24 signals and their complements (for simplicity, only seven input lines are shown). Multiplexers 201A are programmed to select one of the 12 input signals from UIM 400 or one of the signals from fast input pins 500B. Multiplexers 201B are programmed to select one of nine feedback signals from macrocell 200 and input signals from UIM 400. Signals from three lines from UIM 400 are provided directly to AND array 210. The AND array 210 in each FFB 200 generates 45 product terms which drive the nine macrocells 200, with five product terms from the AND array 210 being allocated to each macrocell 220. Four of these product terms are ORed together by an OR gate 225 and drive the input of a programmable Dtype flipflop 222. The fifth product term 223 drives the asynchronous ActiveHigh Set Input to flipflop 222. The clock signal provided to flipflop 222 is one of two global FastCLK signals (FCLK0 or FCLK1) that are distributed with short delay and minimal skew over the entire chip.
Macrocell 220 applies output signals to a fast output pin 500C through a 3state output buffer 224. Each output buffer 224 is controlled by one of two dedicated Fast Output Enable inputs, enabled permanently or disabled permanently. The macrocell output is also routed back as an input to the UIM 400 and as a selectable input to AND array 210.
FIG. 2(B) shows a simplified wiring diagram of an "export chain" (product term assignment) scheme that provides productterm flexibility without disabling macrocell outputs. Product term assignment transfers sumofproduct terms from the output of an OR gate 225 of macrocell 220A to the input terminal of an OR gate 225 of macrocell 220B. This configuration allows complex logic functions having up to 36 product terms to be implemented in one FFB 200. When product terms are assigned to adjacent macrocells 220, the fifth private Pterm 223 may be applied to the Dinput to the macrocell register 222B (See discussion regarding merging during FFB mapping, below). Thus, macrocell register 222B is still usable even though the OR gate 225B is in use.
HighDensity Function Blocks (HDFBs)
FIG. 3 shows a schematic diagram of an HDFB 300. Each HDFB 300 includes an AND array 310 and nine macrocells (one macrocell 320 shown). The AND array 310 of HDFB 300 receives 21 signals and their complements from UIM 400. The AND array 310 provides 5 private product terms per macrocell 320, and also provides 12 sharable product terms, which are uncommitted product terms available for use by any of the nine macrocells within HDFB 300.
Each macrocell 320 includes an Arithmetic Logic Unit (ALU) 321, a flipflop 322, OR gates 323 and 324, and a tristate output buffer 325. The five private product terms received by macrocell 320 are used as the primary input signals for combinatorial functions implemented in ALU 321, or as individual Reset, Set, OutputEnable, and Clock logic functions for flipflop 322. OR gate 323 receives up to four private product terms and up to four shared product terms, whereas OR gate 324 receives the fifth private product term and up to eight of the remaining shared product terms. The output terminal of OR gate 323 is connected to the D1 input of ALU 321. The D2 input terminal is driven by the output signal of OR gate 324.
ALU 321 has two programmable modes: logic and arithmetic. In a logic mode, ALU 321 functions as a 2input function generator using a 4bit lookup table that programmably generates any Boolean function of its D1 and D2 input signals of each macrocell 320. In an arithmetic mode, ALU 321 programmably generates the arithmetic sum or difference of the D1 and D2 input signals. Combined with the carry input signal from an adjacent macrocell 300, ALU 320 operates as a 1bit full adder and generates a carry output to an adjacent macrocell 300.
The output signal of ALU 321 drives the Dinput terminal of flipflop 322. Flipflop 322 is triggered by the clock input, thereby making the Q output signal identical to the D input signal. The clock signal of flip flop 322 is either one of the private product terms or one of two global FastCLK signals (FCLK0 and FCLK1). Global FastCLK signals are distributed to every flipflop 322 with short delay and minimal skew.
In addition to driving output buffer 325, the output signal of macrocell 320 is routed back as an input signal to UIM 400. In this embodiment, either a private product term from macrocell 320 or a global fast OE signal triggers the Output Enable signal (OE Control) to be provided to output buffer 325, thereby providing a feedback signal to UIM 400.
Universal Interconnect Matrix (UIM)
As mentioned previously, UIM 400 receives input signals from macrocell feedback lines, I/O pins 500A, and dedicated input pins 500B (FIG. 1). Acting as an unrestricted crossbar switch, UIM 400 provides 21 output signals for each HDFB 300 and 24 output signals to each FFB 200. In accordance with one embodiment of the present invention, any UIM input line is programmably connectable to any UIM output line. Further, when multiple input lines are programmed to be connected to the same output line, this output line produces the logical AND of the signals on the input lines. By choosing the appropriate signal inversions at the input pins, a macrocell outputs an ANDarray input signal, wherein that AND logic can also be used to implement wide NAND, OR or NOR functions.
Input/Output (I/O) Blocks
FIG. 4 is a simplified schematic diagram showing an I/O block 510 associated with I/O pins 500A and input pins 5005. Output lines from macrocells 320 are connected directly through a 3state output buffer 325 to output pins 500A, each buffer 325 being controlled by an Output Enable product term. Specifically, two dedicated Fast OE input signals, i.e. Fast OE0 and Fast OE1, in conjunction with, the individual Output Enable product term, OE PTerm, control output buffer 325. Each input signal applied to I/O pin 500A is connected to I/O block 510 which can be configured as direct, latched, or registered.
Programming Method
An EPLD programming method according to the present invention will now be discussed.
System Context of the Invention
FIG. 5 shows a "machine" for programming an EPLD 100 in accordance with the present invention. A human user 521 provides the EPLD 100 to be programmed and specifies the logic function (circuit design) 526 which EPLD 100 is to implement. CAD program 524 and logic design software 522, operating in a computer 523, take logic function 526 and determine how to "map" the logic efficiently onto EPLD 100. Specifically, if the specified logic function is represented in a schematic or highlevel language format, logic design software 522 transforms the function from that format to a set of Boolean sumofproducts equations prior to carrying out the steps in FIGS. 611 (described in detail below). The set of Boolean equations includes both sequential (registered) equations and combinatorial equations. Such format transformation procedures are well known in the art and readily available.
CAD software 524 is used after logic design software 522, and produces a bitmap file that indicates to a device programmer 525 the values to program into EPLD 100. This bitmap file, also known as a hex file, is a list of the programmable connections of the ANDarray and ORarray of each function block, of the logic expander or interconnect matrix, and other settings of EPLD 100.
Device programmer 525 physically programs the contents of the bitmap file into EPLD 100. The physical form of the programming or configuring depends on the manufacturing technology of the EPLD 100. For example, if EPLD 100 is an EPROM or EEPROM device, the EPROM or EEPROM cells are programmed by charging or discharging a floating gate or other capacitance element. Other PLD devices are programmed by blowing fuses. One typical device programmer 25 is an HW120 programmer from Xilinx, Inc. This programmer connects to computer 23 via an RS232 serial port. A software driver, such as a Prolink(TM) driver, also available from Xilinx, Inc., downloads programs and data from computer 23 to device programmer 25. Thus, in this manner, EPLD 100 is reduced from an unprogrammed state to a programmed state that carries out a user specified logic function.
Major ComponentsOverview
FIG. 6 is a block diagram showing the main subroutines (procedures) embodying the programming method of the present invention. The program 600 sequentially executes a logic optimization subroutine 610, a logic splitting subroutine 620, a UIM ANDing subroutine 630, an FFB Mapping subroutine 640, an HDFB mapping subroutine 650, a refinement subroutine 660, and a bitmapping (downloading) procedure. Each of these subroutines will be described in greater detail in the following sections.
Logic Optimization Subroutine (FIG. 7)
The first subroutine called by the program is the Logic Optimization subroutine 610 which reformulates the original Boolean equations of logic function 526 into "optimized" Boolean equations likely to map into the function blocks of EPLD 100.
Logic optimization subroutine 610 determines in step 701 whether any of the equations of the logic function are designated, or "marked" by user 521 during design entry into computer 523 as being "fast" equations (i.e., implementable in the FFBs). If "fast" equations are designated, then the subroutine proceeds to step 702; otherwise, the subroutine proceeds to step 710. In step 702, the subroutine determines whether sufficient Fast Clock resources are actually available to support all of the designated "fast" equations. As described previously, there are only two global Fast Clocks which are used to clock registers 222 of FFBs 220 (FIG. 2(A)).
If the "fast" equations require three or more unique Fast Clock signals, then the logic equation cannot implement the logic function because three different clock signals cannot be transmitted on the two Fast Clock lines. When this occurs, the subroutine branches to step 703 which indicates an error on a computer screen (or other output device) and terminates the program. However, if the "fast" equations require two or less Fast Clock lines, then the program proceeds to step 704.
In step 704, a second test is performed to determine if the "fast" equations require more than two Fast Output Enable resources. Similar to the Fast Clock lines, each EPLD 100 of the XC7300 EPLD family has only two global Fast Output Enable lines controlling output buffer 224 (FIG. 2(A)). If the "fast" equations require three or more unique Fast Output Enable signals, then the logic equation cannot implement the logic function. When this occurs, the subroutine branches to step 703 which indicates an error and terminates the program. However, if the "fast" equations require two or less Fast Output Enable lines, then the program proceeds to step 710.
In step 710, logic optimization subroutine 610 identifies registered equations (as opposed to combinatorial equations) which are implementable by the latches/flip flops of input block 510 (see FIG. 4). By implementing registered inputs in the input blocks, macrocell flip flops resources are preserved for other functions. The process of determining whether a registered equation (REQN) is implementable in an input register is determined in accordance with the following pseudocode procedure:
______________________________________Begin procedure;For each REQN in the logic function do:If (REQN input is a single input signal)and (REQN input does not drive internal logic),thenif REQN clock is a previouslydefined Fast Clock,then implement REQN in an input block,else designate REQN as FFBOK;End procedure.______________________________________
The registered equations designated as "FFBOK" are then analyzed to determine if they can be efficiently implemented in an input block. This is determined by the most efficient use of Fast Clock lines. In addition to the two Fast Clock lines connected to the flipflops of the function blocks, the input blocks 510 receive a third Fast Clock line (FastCLK2 in FIG. 4). The subroutine determines the registered equation whose clock signal produces the most efficient use of the third Fast Clock Line. This use is determined by assigning the FFBOK registered equation a weight value according to the following function:
Weight (REQN)=c1*a+c2*bc3*c
where:
c_{1}, c_{2}, c_{3} are constants, and
a=largest possible number of input registers implemented by the logic function multiplied by an optimization constant (one whose value is selected to optimize results);
b=largest possible number of FFBs implemented by the logic function multiplied by an optimization constant; and
c=number of FFBs prevented from implemented by the logic function multiplied by an optimization constant.
After weight values are determined for all of the FFBOK registered equations, then the equation with the largest weight is designated as an input register and that equation's clock signal is defined as the Fast Clock signal. Any other qualified equations using the same clock signal are also implemented in input blocks. This process is then repeated until all qualified registered equations are implemented in input blocks or until all three Fast Clock lines are defined.
Next, in step 720, the logic optimization subroutine 610 determines whether any of the two global Fast Clock lines (those applied only to the function blocks) are unused. That is, if the clock signals for the designated "fast" equations and the input block equations are designated as "don't care" by the user (i.e. not used explicitly), then both Fast Clock lines are free for use in either the HDFBs or the FFBs. Similarly, if only one of the two global Fast Clock lines is used by the "fast" equations, then the second Fast Clock line is available for use in the HDFBs. If neither of the lines are available, then subroutine 610 proceeds to step 730. However, if one or both of the lines are available, then subroutine 610 proceeds to step 725.
In step 725, subroutine 610 determines the most efficient use of the available Fast Clock lines by identifying the most common clock signals associated with the registered equations. This process begins by identifying all registered equations whose clock signals are not previouslydefined Fast Clock signals. Subroutine 610 then assigns a weighted value to each clock signal which is determined by the global effect of defining the equation's clock signal as a Fast Clock. That is, the clock signal of one registered equation is tentatively designated as a Fast Clock. The subroutine determines the clock signal that produces the most efficient use of the third Fast Clock Line. This use is determined by assigning the clock signal a weight value according to the following function:
Weight (FCLK)=c1*a+c2*bc3*c+c4*d
where:
c_{1}, c_{2}, c_{3}, c_{y} are constants
Among the tentatively designated Fast Clock
The subroutine determines the clock signal that produces the most efficient use of an available Fast Clock. This is determined by assigning a weight value to each tentatively marked fast clock according to the following function:
a=number of Pterm clocks eliminated by the tentative assignment;
b=number of FFBs used as a result of the tentative assignment;
c=number of FFBs prevented from implemented by the tentative assignment; and
d=penalty for using a macrocell to create a Fast Clock
After weight values are determined for all of the clock signals, then the clock signal associated with the equation having the largest weight is designated as a Fast Clock signal. This process is then repeated for the second Fast Clock line, if it is available. The subroutine then proceeds to step 730.
In step 730, logic optimization subroutine 610 determines whether any of the two Fast Output Enable lines are unused by the designated "fast" equations. That is, if the output enable signals for the designated "fast" equations are designated as "don't care" by the user, then both Fast Output Enable lines are free for use in either the HDFBs or the FFBs. Similarly, if only one of the Fast Output Enable lines is used by the "fast" equations, then the second Fast Output Enable line is available for use in the HDFBs. If neither of the Fast Enable lines are available, then the subroutine proceeds to step 740. However, if one or both of the lines are available, then the subroutine proceeds to step 735.
In step 735, subroutine 610 determines the most efficient use Of the available Fast Output Enable lines by identifying the most common enable signals associated with the equations. This process begins by identifying all equations whose enable signals are not previouslydefined Fast Output Enable signals. The Logic Optimization subroutine then assigns a weighted value to each equation which is determined by the effect of defining the equation's enable signal as a Fast Output Enable. That is, the enable signal of one equation is tentatively designated as a Fast Output Enable. The subroutine determines the enable signal that produces the most efficient use of the third Fast Output Enable. This use is determined by assigning the enable signal a weight value according to the following function:
Weight (FOE)=c1*a+c2*bc3*c+c4*d
where:
c_{1}, c_{2}, c_{3}, c_{y} are constants
a=number of Pterm clocks eliminated by the tentative assignment;
b=number of FFBs used as a result of the tentative assignment;
c=number of FFBs prevented from implemented by the tentative assignment; and
d=penalty for using a macrocell to create a Fast Output Enable
After weight values are determined for all of the enable signals, the signals are sorted according to their weight value, and the enable signal associated with the equation having the highest weight is designated as a Fast Output Enable Signal. If both Fast Output Enable lines are available, the process is repeated for the second line. Subroutine 610 then proceeds with step 740.
Step 740 includes logic collapsing, which is a wellknown process of combining two or more related equations into larger equations for more efficient implementation in the function blocks. According to one embodiment of the present invention, collapsing is controlled by the Pterm and input signal limits associated with a desired function block type. For example, with respect to the XC7300 EPLD family, FFB macrocells receive 4 Pterms, whereas HDFB macrocells receive up to 17 Pterms. Therefore, collapsing takes into account whether a particular combination forces two FFBOK equations into a condition in which the combined equation may only be implemented in an HDFB. If so, then the combination is prevented.
Finally, in step 750 logic minimization is performed using a wellknown method such as that of the Espresso (a logic minimizer developed at the University of California, Berkeley). The Logic Optimization subroutine then terminates and the program continues with logic splitting subroutine 620 (see FIG. 6).
Logic Splitting Subroutine (FIG. 8)
Logic splitting subroutine 620 splits large logic portions which cannot be implemented (mapped) directly into an FFB or HDFB into two or more subportions. For example, the HDFB of the XC7300 EPLD family includes 21 input terminals and nine output terminals connected to the UIM. Thus, if a logic portion includes more than 21 input signals or more than nine output signals, then that logic portion must be split into two or more subportions to be implemented by the EPLD. In accordance with the present invention, the splitting subroutine divides logic to fit within each of the multiple types of logic blocks. This division requires consideration of the specific circuitry associated with each type of logic block.
With respect to the XC7300 EPLD family, there are two forms of splitting: FFB (or export chain) splitting using the export chain feature of the FFBs 200 (discussed above), and HDFB splitting using the AND gate feature of the UIM 400. Referring back to FIG. 2(B), FFB splitting involves dividing a Boolean equation having more than four product terms into two or more subequations, each having four or fewer product terms. For example, the Boolean equation:
/y=P.sub.1 + . . . +P.sub.8
is split into the two subequations:
/y1=P.sub.1 + . . . +P.sub.4
/y2=P.sub.5 + . . . +P.sub.8.
Referring to FIG. 2(B), /y1 is then assigned to the macrocell 220A, and /y2 is assigned to macrocell 220B. Next, the output signal of OR gate 225A is "exported" via the product term assignment line 226 to the specialized fifth input terminal of OR gate 225B of macrocell 220B.
The HDFB splitting form uses UIM 400 to AND together output signals from the macrocells of FFBs 200 and HDFBs 300. For example, the Boolean equation:
y=P.sub.1 + . . . +P.sub.30
can be split into two subequations:
/y1=P.sub.1 + . . . +P.sub.15
/y2=P.sub.16 + . . . +P.sub.30.
third equation:
/y=/y1*/y2
is produced which is implemented in one embodiment in UIM 400 using the UIMANDing subroutine discussed below.
FIG. 8 shows a flow diagram illustrating the essential portions of logic splitting subroutine 620 which includes both FFB and HDFB splitting. Upon entering logic splitting subroutine 620, all of the equations of the logic function are stored in a file (not shown for simplicity). In step 801, one of the stored equations is selected from the file. Step 810, determines whether that equation is designated as "fast" (i.e., an FFBdesignated equation). If the equation is an FFBdesignated equation, subroutine 620 branches to step 811; otherwise, the subroutine continues with step 820.
In step 811, the FFBdesignated equation is checked to determine whether the equation fits within the parameters of the FFB. In other words, if the designated equation is determined to be too large for one FFB then in step 812, an error message is provided on an output device indicating a fail error has occurred in the programming process, and the program is terminated. However, if the selected equation meets the criteria for mapping into an FFB, subroutine 620 continues with step 813. In step 813, subroutine 620 determines whether the selected equation contains less than or equal to 4 Pterms. If the equation has less than or equal to 4 Pterms, then splitting is unnecessary and the equation is designated as FFBOK in step 815. However, if the equation has more than four Pterms, the subroutine continues with step 814, in which FFB splitting is performed using the export chain feature of the FFBs (discussed in detail above). Then, the equation is designated as FFBOK in step 815, and the subroutine defaults to step 860 (discussed below).
However, if an equation is not FBBdesignated in step 810, the number of signals associated with the selected equation are then counted to determine whether there are greater than 21 but less than or equal to 24 input signals in step 820. If this test is met, subroutine 620 transfers to step 821 in which a multipart test determines if the nondesignated equation is implementable in an FFB. Specifically, the test requires that (1) the equation is used in negative polarity and has less than or equal to 4 Pterms; (2) and the equation does not use a Pterm output enable signal, a Pterm clock signal and is not used in negative polarity in a UIM AND function; (3) the total number of FFBOK equations is less than or equal to the total number of FFB macrocells minus 1, and (4) the number of different inputs used in all FFB equations including the selected equation is less than or equal to the total number of input signals to the FFBs. If all four elements of this test are met, then subroutine 620 transfers to step 822 in which the selected equation is designated as FFBOK, and the subroutine then defaults to step 860. If the outcome of this test is negative, subroutine 610 passes to step 830.
In step 830, the selected equation is tested to determine if the number of input signals is less than or equal to 21 and the number of Pterms is less than or equal to 17, thereby determining whether the equation fits within one HDFB macrocell. If the outcome of this test is positive, then subroutine 620 branches to step 860. If the outcome of this test is negative, the subroutine continues with step 840.
In step 840, the selected equation is divided into two or more subequations having less than or equal to 21 input signals and less than or equal to 17 Pterms. In one embodiment, subequations are formed which have 8 or fewer Pterms. However, note this number is reduced during Refinement Subroutine 660 (described below). In step 850, a new equation is formed which is a sum of products equation combining the outcomes of all subexpressions divided from the selected equation. This equation is designated as "UIMOK". Subroutine 620 then defaults to step 860.
In step 860, the subroutine checks to determine if all equations have been processed. If any equations are unprocessed, subroutine 620 returns to step 801; otherwise subroutine 620 ends and control is passed to the UIMANDing subroutine 830.
UIM ANDing Subroutine (FIGS. 9A and 9B)
UIM ANDing is used as an alternative to placing an equation in a macrocell. Using the UIM to form AND gates typically results in significant advantages, such as requiring fewer macrocells and fewer passes of logic, thereby creating a faster circuit. However, a dualblock architecture has the disadvantage of limiting the equations which are eligible to be placed in FFBs. For this reason, care must be taken in deciding which equations are formed using the UIM AND.
The UIMAND array can form an AND function using any of the input signals into the UIM. For instance, in one embodiment an internal node of the form n=a*b*c is formed in the UIM. Conversely, if the internal node is equal to n=a+b+c, an equation n=/a*/b*/c is created using DeMorgan's theorem, and the inverter is pushed ahead into a subsequent equation to which n is fed.
"Polarity" problems are created when output signals from FFB macrocells are fed back into the UIM. Because these feedback signals are applied on single lines (there is no complementary line) the polarity of these signals is fixed at the output line of the FFB. Therefore, a polarity problem can arise if the signals from the FFB used to form a UIM AND equation are of an incorrect polarity.
In contrast, output signals from the HDFBs are also fed back into the UIM on single lines, but these signals are inverted by the ALU, if necessary. Specifically, the ALU includes a lookup table which generates any logic function of two inputs. Therefore, any designated equation can be inverted by the ALU to provide a desired polarity at the output of the HDFB's macrocell.
FIGS. 9(A) and 9(B) show essential steps associated with UIM ANDing subroutine 630. Referring to FIG. 9(A), UIM ANDing subroutine 630 begins by determining whether any of the equations associated with the logic function qualify as "UIMOK". The process for determining whether an equation is UIMOK is illustrated in FIG. 9(B). As shown in FIG. 9B, step 911 determines whether an equation qualifies as a "node". The term node is used to designate an equation whose outputs are not applied to an EPLD I/O pin. In other words, all output signals of a node are received as input signals in a subsequent equation. Further, at least one node of the logic equation must meet the following multistep test:
Step 912the node cannot be registered.
Step 913the node must be expressed as a single Pterm.
Step 914the node must meet the following polarity requirements:
(a) all node input signals must not be used in another UIMOK equation in an opposite polarity and (b) all node input signals which are feedback signals from an FFB macrocell must have a positive polarity (since the device does not provide inversion resources for feedback signals from FFBs).
Step 915the node, if implemented as a UIMAND function, must not cause an undesirable number of FFBOK equations to lose their FFBOK status (i.e., require these nodes to be implemented only in an HDFB). In one embodiment, if the number of FFBOK equations falls below 18, then the node is not implemented as a UIMAND function.
If all of steps 911915 are satisfied, then the node is considered UIMOK in step 916. Otherwise, the node is designated not UIMOK in step 917. This process is then performed on all equations of the logic function, and if one or more of the equations are UIMOK, then a positive response is returned in step 910.
If a UIMOK node exists in step 910, UIM ANDing subroutine 630 continues with step 920 in which a weight value is calculated for each UIMOK node. Weights are determined based on the following equation:
Weight (Node/EQN)=c1c2c3
where:
c1=number of inputs to node;
c2=number of nodes which lose UIMOK status if the node is implemented as a UIMAND equation; and
c3=number of equations which lose FFBOK status if the node is implemented as a UIMAND equation.
After all nodes are provided with weight values in step 920, the node having the highest weight value is designated as a UIMAND function in step 930. By designating a node as a UIMAND function, other equations in the logic function may become UIMOK and others may lose their UIMOK status. Therefore, after step 930, subroutine 630 loops up to step 910, and this process is continued until no UIMOK nodes exist in the logic function. When no such nodes exist, subroutine 630 ends.
FFB Mapping Subroutine (FIG. 10)
The purpose of FFB mapping subroutine 640 is to use as many resources of the FFBs as possible before resorting to mapping on the HDFBs. FFB mapping subroutine 640 begins by mapping all of the userdefined "fast" logic portions. After all of the "fast" logic portions have been mapped, FFB mapping subroutine 640 then maps undesignated logic portions into the FFBs. The subroutine may transform undesignated logic portions to fit within an FFB (e.g., inverts the function by changing polarities everywhere else in the logic portion).
FIG. 10 shows a flow chart of FFB mapping subroutine 640. In step 1010, a list is formed of all FFB resources. Next, in step 1020, all of the equations which are designated as FFBOK are provided a weight value and sorted from highest weight value to lowest weight value. An equation is designated FFBOK if it meets all of the following tests:
(1) The block doesn't require a Pterm Output Enable signal or a Pterm Clock signal.
(2) The block doesn't explicitly use the HDFB ALU circuit.
(3) The block doesn't use a flipflop reset signal.
(4) The block doesn't include more than 4 pterm in negative polarity.
The weight value is determined by userdesignated FFB equations, which are given the highest weight, and other FFBOK equations which are weighted by the number of inputs in the equation. In step 1030, subroutine 640 determines if any FFBs have been unmapped. Logically, when subroutine 640 is initiated there are FFBs left unmapped. Therefore, subroutine 640 typically passes to step 1040 in which the FFB having a lowest weight is selected and then removed from the list. Following this, in step 1050 a list is formed of all unmapped FFBOK equations (of course, the first pass through this step yields a list of all equations weighted in step 1020). In step 1060, subroutine 640 determines whether the equation having the highest weight value can be mapped into the selected FFB. If so, in step 1061 the equation is mapped into the s FFB, and in step 1062 the mapped equation is removed from the list. However, if the highest weighted equation cannot be mapped into the FFB, the equation is removed from the list into step 1062. Step 1063 determines whether any equations are left in a list. If so, subroutine 640 loops up to step 1060 wherein steps 10601063 are repeated until a negative outcome is determined in step 1063.
Following step 1063, a merging process is performed in step 1070 wherein single Pterm equations are assigned to macrocells of the FFB in which the export chain is being used (as discussed above).
Finally, in step 1080, a single function block refinement process is performed in which a list of weighted, unplaced FFBOK equations is formed, and previously mapped equations are temporarily removed from the FFB. If two or more of the unplaced equations can be mapped into the FFB in place of the previously mapped equation, then the previously mapped equation is "unmapped", and the two or more equations are mapped into the FFB.
Following step 1080 the mapping subroutine loops back to step 1030 which determines whether any FFBs are left to be mapped. After all FFBs have been mapped, the outcome of step 1030 is negative and subroutine 640 ends.
FDFB Mapping Subroutine (FIG. 11)
The HDFB Mapping subroutine 650 attempts to fill each HDFB as full as possible. The steps to accomplish HDFB mapping (as shown in FIG. 11) are similar to FFB mapping (as shown in FIG. 10) with four exceptions. Note that similar steps in the two subroutines have identical reference numbers in the last two digits.
First exception: in step 1110, each HDFB is given a weight based on the number of I/O pins 500A connected to the macrocells of the HDFB.
Second exception: in step 1125, the termination decision is determined by whether any equation of the logic function have been left unmapped. Once all equations have been mapped, HDFB mapping subroutine 650 ends.
Third exception: in step 1135, if a negative outcome occurs in step 1030, then all HDFBs of an EPLD have been mapped and there are left over equations. In this case, a "virtual" HDFB is formed in the computer memory which is assigned the same number of macrocells, input signals and output signals as the highest weighted HDFB. This virtual HDFB then becomes the selected HDFB for the mapping procedure, and the subroutine proceeds to step 1150.
Fourth exception: in step 1070 (FIG. 10) FFB mapping is not performed in HDFB mapping subroutine 650 because the HDFB does not include the export chain feature found in the FFBs.
Refinement Subroutine (FIG. 12)
Refinement Subroutine 660 analyzes the results of FFB mapping subroutine 640 and HDFB mapping subroutine 650 and tries to make corrective actions, if necessary. If the FFBs are underutilized, Refinement Subroutine 660 checks to see if any remaining FFBOK equations can be mapped into the FFBs. Similarly, if any equations are unmapped, the Refinement Subroutine 660 splits large equations to make more efficient use of the HDFBs.
FIG. 12 shows a flow diagram illustrating the essential steps of Refinement Subroutine 660. In step 1201, a list is created including all FFBs of the target device. In step 1210, the subroutine determines whether all FFBs have been considered for refinement. Of course, when step 1210 is first executed, the result of this decision is always negative.
Assuming a negative result in step 1210, in step 1211 the subroutine selects one FFB from the FFB list (created in step 1201). In step 1212, the subroutine then determines whether the selected FFB contains two or more unused (unmapped) macrocells. If the outcome of step 1212 is negative, then the subroutine returns to step 1210. If the outcome of step 1212 is positive, then the subroutine proceeds with step 1213.
In step 1213, a second list is created containing all FFBOK equations which are either mapped in an actual HDFB or mapped in a virtual HDFB. In determining whether these equations are FFBOK, each equation is subjected to tests similar to those described above in the FFB Mapping subroutine. In addition, if the equation contains more than four Pterms (i.e., export chain splitting is required), then the selected FFB is checked to determine if the unmapped macrocells are adjacent (linked by an export chain). If the equation is too large for the unmapped macrocells, or if the unmapped macrocells are not adjacent, then the equation is not designated as FFBOK. The FFBOK equations are then sorted using a weight value determined as follows:
Weight (EQN)=c1+c2+c3
where:
c1 is inversely proportional to the number of macrocells required to implement the equation;
c2 is proportional to the number of input signals to the equation; and
c3 is proportional to the number of Pterms contained in the equation.
Next, in step 1214 the equation having the largest weight value is selected from the list. In step 1215, export chain splitting is performed on the equation and, in step 1216, the equation is tested for mapping into the FFB. If the equation can be mapped using export chain splitting, then mapping is performed in step 1217. Following a negative result in step 1216 or after mapping in step 1217, the subroutine proceeds to step 1218.
In step 1218, the selected FFB is checked to determine if any unmapped macrocells still exist. If not, the subroutine returns to step 1210. If there are unmapped macrocells, then the subroutine proceeds to step 1219 which determines whether there are any equations remaining in the list (created in step 1213). If the list contains additional equations, then the subroutine returns to step 1214 wherein the nexthighest weighted equation is selected. If the list is empty, then the subroutine returns to step 1210, which checks whether any FFBs remain in the list created in step 1201.
After all the FFBs are considered for refinement, the Refinement subroutine then proceeds to step 1220. In step 1220, the subroutine determines whether any equations remain unmapped after the FFB refinement. If all of the equations are mapped prior to step 1220, then the program ends. If any equations remain unmapped in step 1220, then the subroutine executes steps 1221 through 1228.
In step 1221, a list is created containing all the unmapped equations. In addition, equations are extracted (unmapped) from HDFBs having seven or fewer mapped macrocells. The purpose of unmapping these HDFBs is to eliminate underutilized HDFBs to provide resources for mapping all of the equations of the logic function. Further, in step 1221, all unmapped macrocells of all of the HDFBs are counted. Next, all of the equations in the file are weighed and sorted. As in other subroutines, the weight assigned to each equation is proportional to its number of inputs signals and number of Pterms. The subroutine then proceeds to step 1223 in which the equation having the highest weight value is selected from the list. In step 1224, the subroutine determines whether splitting the equation into subequations having 4 or fewer Pterms would cause the total number of equations in the list to exceed the number of unmapped macrocells. If the result of step 1224 is positive, then the subroutine defaults to step 1226; otherwise, the selected equation is split in step 1225. In step 1226, the subroutine then determines whether all equations in the list have been considered and, if not, the subroutine returns to step 1223. If all equations have been considered in step 1224, then the subroutine proceeds to step 1227 in which all of the HDFBs are unmapped. Finally, in step 1228 the subroutine calls the HDFB Mapping subroutine 650 (discussed above), and then quits.
Conclusion
Although the present invention has been described in considerable detail with reference to certain preferred embodiments thereof, other embodiments are possible. For example, the function block types can contain lookup tables instead of ANDOR arrays. In addition, the execution order of some of the described subroutines can be rearranged without changing the results. Therefore, the spirit and scope of the appended claims should not be limited to the description of the preferred embodiments contained herein. The present invention is set forth in the appended claims.
Claims (22)
Priority Applications (1)
Application Number  Priority Date  Filing Date  Title 

US08370768 US5636368A (en)  19941223  19941223  Method for programming complex PLD having more than one function block type 
Applications Claiming Priority (2)
Application Number  Priority Date  Filing Date  Title 

US08370768 US5636368A (en)  19941223  19941223  Method for programming complex PLD having more than one function block type 
US08818138 US5963048A (en)  19941223  19970312  Method for programming complex PLD having more than one function block type 
Related Child Applications (1)
Application Number  Title  Priority Date  Filing Date 

US08818138 Continuation US5963048A (en)  19941223  19970312  Method for programming complex PLD having more than one function block type 
Publications (1)
Publication Number  Publication Date 

US5636368A true US5636368A (en)  19970603 
Family
ID=23461091
Family Applications (2)
Application Number  Title  Priority Date  Filing Date 

US08370768 Expired  Lifetime US5636368A (en)  19941223  19941223  Method for programming complex PLD having more than one function block type 
US08818138 Expired  Lifetime US5963048A (en)  19941223  19970312  Method for programming complex PLD having more than one function block type 
Family Applications After (1)
Application Number  Title  Priority Date  Filing Date 

US08818138 Expired  Lifetime US5963048A (en)  19941223  19970312  Method for programming complex PLD having more than one function block type 
Country Status (1)
Country  Link 

US (2)  US5636368A (en) 
Cited By (82)
Publication number  Priority date  Publication date  Assignee  Title 

US5790882A (en) *  19961113  19980804  Xilinx, Inc.  Programmable logic device placement method utilizing weighting function to facilitate pin locking 
US5926035A (en) *  19960626  19990720  Cypress Semiconductor Corp.  Method and apparatus to generate mask programmable device 
US5943488A (en) *  19960626  19990824  Cypress Semiconductor Corp.  Method and apparatus to generate mask programmable device 
US5963048A (en) *  19941223  19991005  Xilinx, Inc.  Method for programming complex PLD having more than one function block type 
US6091892A (en) *  19961113  20000718  Xilinx, Inc.  Method for mapping product terms in a complex programmable logic device 
US6144220A (en) *  19960903  20001107  Xilinx, Inc.  FPGA Architecture using multiplexers that incorporate a logic gate 
US6195788B1 (en) *  19971017  20010227  Altera Corporation  Mapping heterogeneous logic elements in a programmable logic device 
US6453456B1 (en) *  20000322  20020917  Xilinx, Inc.  System and method for interactive implementation and testing of logic cores on a programmable logic device 
US6725441B1 (en)  20000322  20040420  Xilinx, Inc.  Method and apparatus for defining and modifying connections between logic cores implemented on programmable logic devices 
US20040199700A1 (en) *  20030403  20041007  Clayton Shawn Adam  Virtual peripheral component interconnect multiplefunction device 
US7119576B1 (en)  20000918  20061010  Altera Corporation  Devices and methods with programmable logic and digital signal processing regions 
US20060247909A1 (en) *  20050427  20061102  Desai Madhav P  System and method for emulating a logic circuit design using programmable logic devices 
US7249329B1 (en) *  20040601  20070724  Altera Corporation  Technology mapping techniques for incomplete lookup tables 
CN1329806C (en) *  20040818  20070801  å“ˆå°”æ»¨å·¥ä¸šå¤§å¦  Highspeed largecapacity data collecting system based on CPLD and SDRAM 
US7346644B1 (en)  20000918  20080318  Altera Corporation  Devices and methods with programmable logic and digital signal processing regions 
US7350176B1 (en) *  20030717  20080325  Altera Corporation  Techniques for mapping to a shared lookup table mask 
US20080133627A1 (en) *  20061205  20080605  Altera Corporation  Large multiplier for programmable logic device 
US20100208194A1 (en) *  20090213  20100819  Amitava Gupta  Variable focus liquid filled lens apparatus 
US20100228806A1 (en) *  20090303  20100909  Keone Streicher  Modular digital signal processing circuitry with optionally usable, dedicated connections between modules of the circuitry 
US20100228807A1 (en) *  20090303  20100909  Altera Corporation  Digital signal processing circuitry with redundancy and bidirectional data paths 
US7814137B1 (en)  20070109  20101012  Altera Corporation  Combined interpolation and decimation filter for programmable logic device 
US7822799B1 (en)  20060626  20101026  Altera Corporation  Adderrounder circuitry for specialized processing block in programmable logic device 
US7836117B1 (en)  20060407  20101116  Altera Corporation  Specialized processing block for programmable logic device 
US7865541B1 (en)  20070122  20110104  Altera Corporation  Configuring floating point operations in a programmable logic device 
US20110085131A1 (en) *  20091014  20110414  Adlens Beacon, Inc.  Aspheric Fluid Filled Lens Optic 
US20110085243A1 (en) *  20091013  20110414  Amitava Gupta  NonRound Fluid Filled Lens Optic 
US20110102735A1 (en) *  20091015  20110505  Amitava Gupta  Fluid Filled Lens Reservoir System and Manufacturing Method of the Reservoir System 
US7948267B1 (en)  20100209  20110524  Altera Corporation  Efficient rounding circuits and methods in configurable integrated circuit devices 
US7949699B1 (en)  20070830  20110524  Altera Corporation  Implementation of decimation filter in integrated circuit device using rambased data storage 
US8041759B1 (en)  20060209  20111018  Altera Corporation  Specialized processing block for programmable logic device 
US8087778B2 (en)  20090213  20120103  Adlens Beacon, Inc.  Variable focus liquid filled lens mechanism 
USD665009S1 (en)  20101014  20120807  Adlens Beacon, Inc.  Spectacles frame 
US8244789B1 (en)  20080314  20120814  Altera Corporation  Normalization of floating point operations in a programmable integrated circuit device 
US8255448B1 (en)  20081002  20120828  Altera Corporation  Implementing division in a programmable integrated circuit device 
US8266198B2 (en)  20060209  20120911  Altera Corporation  Specialized processing block for programmable logic device 
US8266199B2 (en)  20060209  20120911  Altera Corporation  Specialized processing block for programmable logic device 
US8301681B1 (en)  20060209  20121030  Altera Corporation  Specialized processing block for programmable logic device 
US8307023B1 (en)  20081010  20121106  Altera Corporation  DSP block for implementing large multiplier on a programmable integrated circuit device 
US8353593B2 (en)  20091015  20130115  Adlens Beacon, Inc.  Hinge mechanism for a fluid filled lens assembly 
US8386553B1 (en)  20061205  20130226  Altera Corporation  Large multiplier for programmable logic device 
US8386550B1 (en)  20060920  20130226  Altera Corporation  Method for configuring a finite impulse response filter in a programmable logic device 
US8396914B1 (en)  20090911  20130312  Altera Corporation  Matrix decomposition in an integrated circuit device 
US8412756B1 (en)  20090911  20130402  Altera Corporation  Multioperand floating point operations in a programmable integrated circuit device 
US8458243B1 (en)  20100303  20130604  Altera Corporation  Digital signal processing circuit blocks with support for systolic finiteimpulseresponse digital filtering 
US8468192B1 (en)  20090303  20130618  Altera Corporation  Implementing multipliers in a programmable integrated circuit device 
US8484265B1 (en)  20100304  20130709  Altera Corporation  Angular range reduction in an integrated circuit device 
US8488250B2 (en)  20101011  20130716  Adlens Beacon, Inc.  Perimeter piezo reservoir in a lens 
US8510354B1 (en)  20100312  20130813  Altera Corporation  Calculation of trigonometric functions in an integrated circuit device 
US8539014B2 (en)  20100325  20130917  Altera Corporation  Solving linear matrices in an integrated circuit device 
US8539016B1 (en)  20100209  20130917  Altera Corporation  QR decomposition in an integrated circuit device 
US8543634B1 (en)  20120330  20130924  Altera Corporation  Specialized processing block for programmable integrated circuit device 
US8577951B1 (en)  20100819  20131105  Altera Corporation  Matrix operations in an integrated circuit device 
US8589463B2 (en)  20100625  20131119  Altera Corporation  Calculation of trigonometric functions in an integrated circuit device 
US8601044B2 (en)  20100302  20131203  Altera Corporation  Discrete Fourier Transform in an integrated circuit device 
US8620980B1 (en)  20050927  20131231  Altera Corporation  Programmable device with specialized multiplier blocks 
US8626815B1 (en)  20080714  20140107  Altera Corporation  Configuring a programmable integrated circuit device to perform matrix multiplication 
US8645451B2 (en)  20110310  20140204  Altera Corporation  Doubleclocked specialized processing block in an integrated circuit device 
US8645449B1 (en)  20090303  20140204  Altera Corporation  Combined floating point adder and subtractor 
US8645450B1 (en)  20070302  20140204  Altera Corporation  Multiplieraccumulator circuitry and methods 
US8650231B1 (en)  20070122  20140211  Altera Corporation  Configuring floating point operations in a programmable device 
US8650236B1 (en)  20090804  20140211  Altera Corporation  Highrate interpolation or decimation filter in integrated circuit device 
US8706790B1 (en)  20090303  20140422  Altera Corporation  Implementing mixedprecision floatingpoint operations in a programmable integrated circuit device 
US8708486B2 (en)  20091015  20140429  Adlens Beacon, Inc.  Fluid filled lenses and mechanisms of inflation thereof 
US8762443B1 (en)  20111115  20140624  Altera Corporation  Matrix operations in an integrated circuit device 
US8812576B1 (en)  20110912  20140819  Altera Corporation  QR decomposition in an integrated circuit device 
US8817381B2 (en)  20091013  20140826  Adlens Beacon, Inc.  Full field membrane design for nonround liquid lens assemblies 
US8862650B2 (en)  20100625  20141014  Altera Corporation  Calculation of trigonometric functions in an integrated circuit device 
US8886696B1 (en)  20090303  20141111  Altera Corporation  Digital signal processing circuitry with redundancy and ability to support larger multipliers 
US8949298B1 (en)  20110916  20150203  Altera Corporation  Computing floatingpoint polynomials in an integrated circuit device 
US20150046629A1 (en) *  20130809  20150212  Hon Hai Precision Industry Co., Ltd.  Switch apparatus and electronic device 
US8959137B1 (en)  20080220  20150217  Altera Corporation  Implementing large multipliers in a programmable integrated circuit device 
US8996600B1 (en)  20120803  20150331  Altera Corporation  Specialized processing block for implementing floatingpoint multiplier with subnormal operation support 
US9036264B2 (en)  20100812  20150519  Adlens Beacon, Inc.  Fluidfilled lenses and their ophthalmic applications 
US9042027B2 (en)  20101110  20150526  Adlens Beacon, Inc.  Fluidfilled lenses and actuation systems thereof 
US9053045B1 (en)  20110916  20150609  Altera Corporation  Computing floatingpoint polynomials in an integrated circuit device 
US9098332B1 (en)  20120601  20150804  Altera Corporation  Specialized processing block with fixed and floatingpoint structures 
US9189200B1 (en)  20130314  20151117  Altera Corporation  Multipleprecision processing block in a programmable integrated circuit device 
US9207909B1 (en)  20121126  20151208  Altera Corporation  Polynomial calculations optimized for programmable integrated circuit device structures 
US9348795B1 (en)  20130703  20160524  Altera Corporation  Programmable device using fixed and configurable logic to implement floatingpoint rounding 
US9379687B1 (en)  20140114  20160628  Altera Corporation  Pipelined systolic finite impulse response filter 
US9600278B1 (en)  20110509  20170321  Altera Corporation  Programmable device using fixed and configurable logic to implement recursive trees 
US9684488B2 (en)  20150326  20170620  Altera Corporation  Combined adder and preadder for highradix multiplier circuit 
Families Citing this family (37)
Publication number  Priority date  Publication date  Assignee  Title 

US6255847B1 (en) *  19980521  20010703  Lattice Semiconductor Corporation  Programmable logic device 
US6229336B1 (en)  19980521  20010508  Lattice Semiconductor Corporation  Programmable integrated circuit device with slew control and skew control 
US6320410B1 (en) *  19990317  20011120  Cypress Semiconductor Corp.  Heterogeneous CPLD logic blocks 
US6414871B1 (en)  20000512  20020702  Xilinx, Inc.  Systems and methods for programming programmable devices 
CA2338458A1 (en) *  20010227  20010814  Ioan Dancea  Method and vlsi circuits allowing to change dynamically the logical behaviour 
US7752419B1 (en)  20010322  20100706  Qst Holdings, Llc  Method and system for managing hardware resources to implement system functions using an adaptive computing architecture 
US7325123B2 (en) *  20010322  20080129  Qst Holdings, Llc  Hierarchical interconnect for configuring separate interconnects for each group of fixed and diverse computational elements 
US7624204B2 (en) *  20010322  20091124  Nvidia Corporation  Input/output controller node in an adaptable computing environment 
US6836839B2 (en)  20010322  20041228  Quicksilver Technology, Inc.  Adaptive integrated circuitry with heterogeneous and reconfigurable matrices of diverse and adaptive computational units having fixed, application specific computational elements 
US7962716B2 (en)  20010322  20110614  Qst Holdings, Inc.  Adaptive integrated circuitry with heterogeneous and reconfigurable matrices of diverse and adaptive computational units having fixed, application specific computational elements 
US6577678B2 (en)  20010508  20030610  Quicksilver Technology  Method and system for reconfigurable channel coding 
US7046635B2 (en)  20011128  20060516  Quicksilver Technology, Inc.  System for authorizing functionality in adaptable hardware devices 
US8412915B2 (en)  20011130  20130402  Altera Corporation  Apparatus, system and method for configuration of adaptive integrated circuitry having heterogeneous computational elements 
US6986021B2 (en)  20011130  20060110  Quick Silver Technology, Inc.  Apparatus, method, system and executable module for configuration and operation of adaptive integrated circuitry having fixed, application specific computational elements 
US7802108B1 (en)  20020718  20100921  Nvidia Corporation  Secure storage of program code for an embedded system 
US7644279B2 (en) *  20011205  20100105  Nvidia Corporation  Consumer product distribution in the embedded system market 
US7215701B2 (en)  20011212  20070508  Sharad Sambhwani  Low I/O bandwidth method and system for implementing detection and identification of scrambling codes 
US7403981B2 (en)  20020104  20080722  Quicksilver Technology, Inc.  Apparatus and method for adaptive multimedia reception and transmission in communication environments 
US7620678B1 (en)  20020612  20091117  Nvidia Corporation  Method and system for reducing the timetomarket concerns for embedded system design 
US7653710B2 (en)  20020625  20100126  Qst Holdings, Llc.  Hardware task manager 
US7433909B2 (en)  20020625  20081007  Nvidia Corporation  Processing architecture for a reconfigurable arithmetic node 
US8108656B2 (en) *  20020829  20120131  Qst Holdings, Llc  Task definition for specifying resource requirements 
US7502915B2 (en) *  20020930  20090310  Nvidia Corporation  System and method using embedded microprocessor as a node in an adaptable computing machine 
US7937591B1 (en)  20021025  20110503  Qst Holdings, Llc  Method and system for providing a device which can be adapted on an ongoing basis 
US7249242B2 (en)  20021028  20070724  Nvidia Corporation  Input pipeline registers for a node in an adaptive computing engine 
US8949576B2 (en) *  20021101  20150203  Nvidia Corporation  Arithmetic node including general digital signal processing functions for an adaptive computing machine 
US8276135B2 (en)  20021107  20120925  Qst Holdings Llc  Profiling of software and circuit designs utilizing data operation analyses 
US7225301B2 (en)  20021122  20070529  Quicksilver Technologies  External memory controller node 
US7064577B1 (en) *  20030502  20060620  Altera Corporation  Method and apparatus for supporting variable speed configuration hardware 
US7660984B1 (en)  20030513  20100209  Quicksilver Technology  Method and system for achieving individualized protected space in an operating system 
US7328414B1 (en)  20030513  20080205  Qst Holdings, Llc  Method and system for creating and programming an adaptive computing engine 
US8296764B2 (en)  20030814  20121023  Nvidia Corporation  Internal synchronization control for adaptive integrated circuitry 
US7174432B2 (en)  20030819  20070206  Nvidia Corporation  Asynchronous, independent and multiple process shared memory system in an adaptive computing architecture 
US7057412B1 (en) *  20031212  20060606  Altera Corporation  Configurable crossbar switch 
US7129747B1 (en) *  20041015  20061031  Xilinx, Inc.  CPLD with fast logic sharing between function blocks 
US7987065B1 (en)  20070417  20110726  Nvidia Corporation  Automatic quality testing of multimedia rendering by software drivers 
US7759972B1 (en) *  20071031  20100720  Altera Corporation  Integrated circuit architectures with heterogeneous highspeed serial interface circuitry 
Citations (19)
Publication number  Priority date  Publication date  Assignee  Title 

US4791602A (en) *  19830414  19881213  Control Data Corporation  Soft programmable logic array 
US4876466A (en) *  19871120  19891024  Mitsubishi Denki Kabushiki Kaisha  Programmable logic array having a changeable logic structure 
US5015884A (en) *  19850329  19910514  Advanced Micro Devices, Inc.  Multiple array high performance programmable logic device family 
US5034899A (en) *  19860707  19910723  Bbc Brown Boveri Ag  Software tool for automatically generating a functionaldiagram graphic 
US5089973A (en) *  19861107  19920218  Apple Computer Inc.  Programmable logic cell and array 
US5128871A (en) *  19900307  19920707  Advanced Micro Devices, Inc.  Apparatus and method for allocation of resoures in programmable logic devices 
US5165086A (en) *  19850220  19921117  Hitachi, Ltd.  Microprocessor chip using twolevel metal lines technology 
US5202592A (en) *  19891208  19930413  Kawasaki Steel Corporation  Programmable logic device 
US5283875A (en) *  19900413  19940201  Digital Equipment Corporation  Method and apparatus for optimizing prefetch caching by reverse ordering of logical blocks 
US5349250A (en) *  19930902  19940920  Xilinx, Inc.  Logic structure and circuit for fast carry 
US5349691A (en) *  19900703  19940920  Xilinx, Inc.  Programming process for 3level programming logic devices 
US5357152A (en) *  19921110  19941018  Infinite Technology Corporation  Logic system of logic networks with programmable selected functions and programmable operational controls 
US5362999A (en) *  19930318  19941108  Xilinx, Inc.  EPLD chip with hybrid architecture optimized for both speed and flexibility 
US5408628A (en) *  19920707  19950418  Odetics, Inc.  Solid state recorder with flexible width data bus utilizing lock mapping and error correction and detection circuits 
US5440247A (en) *  19930526  19950808  Kaplinsky; Cecil H.  Fast CMOS logic with programmable logic control 
US5448493A (en) *  19891220  19950905  Xilinx, Inc.  Structure and method for manually controlling automatic configuration in an integrated circuit logic block array 
US5457409A (en) *  19920803  19951010  Advanced Micro Devices, Inc.  Architecture of a multiple array high density programmable logic device with a plurality of programmable switch matrices 
US5489857A (en) *  19920803  19960206  Advanced Micro Devices, Inc.  Flexible synchronous/asynchronous cell structure for a high density programmable logic device 
US5499192A (en) *  19911030  19960312  Xilinx, Inc.  Method for generating logic modules from a high level block diagram 
Family Cites Families (1)
Publication number  Priority date  Publication date  Assignee  Title 

US5636368A (en) *  19941223  19970603  Xilinx, Inc.  Method for programming complex PLD having more than one function block type 
Patent Citations (20)
Publication number  Priority date  Publication date  Assignee  Title 

US4791602A (en) *  19830414  19881213  Control Data Corporation  Soft programmable logic array 
US5165086A (en) *  19850220  19921117  Hitachi, Ltd.  Microprocessor chip using twolevel metal lines technology 
US5015884A (en) *  19850329  19910514  Advanced Micro Devices, Inc.  Multiple array high performance programmable logic device family 
US5034899A (en) *  19860707  19910723  Bbc Brown Boveri Ag  Software tool for automatically generating a functionaldiagram graphic 
US5089973A (en) *  19861107  19920218  Apple Computer Inc.  Programmable logic cell and array 
US4876466A (en) *  19871120  19891024  Mitsubishi Denki Kabushiki Kaisha  Programmable logic array having a changeable logic structure 
US5202592A (en) *  19891208  19930413  Kawasaki Steel Corporation  Programmable logic device 
US5448493A (en) *  19891220  19950905  Xilinx, Inc.  Structure and method for manually controlling automatic configuration in an integrated circuit logic block array 
US5128871A (en) *  19900307  19920707  Advanced Micro Devices, Inc.  Apparatus and method for allocation of resoures in programmable logic devices 
US5283875A (en) *  19900413  19940201  Digital Equipment Corporation  Method and apparatus for optimizing prefetch caching by reverse ordering of logical blocks 
US5349691A (en) *  19900703  19940920  Xilinx, Inc.  Programming process for 3level programming logic devices 
US5499192A (en) *  19911030  19960312  Xilinx, Inc.  Method for generating logic modules from a high level block diagram 
US5408628A (en) *  19920707  19950418  Odetics, Inc.  Solid state recorder with flexible width data bus utilizing lock mapping and error correction and detection circuits 
US5457409A (en) *  19920803  19951010  Advanced Micro Devices, Inc.  Architecture of a multiple array high density programmable logic device with a plurality of programmable switch matrices 
US5489857A (en) *  19920803  19960206  Advanced Micro Devices, Inc.  Flexible synchronous/asynchronous cell structure for a high density programmable logic device 
US5357152A (en) *  19921110  19941018  Infinite Technology Corporation  Logic system of logic networks with programmable selected functions and programmable operational controls 
US5450021A (en) *  19930318  19950912  Xilinx, Inc.  EPLD chip with hybrid architecture optimized for both speed and flexibility 
US5362999A (en) *  19930318  19941108  Xilinx, Inc.  EPLD chip with hybrid architecture optimized for both speed and flexibility 
US5440247A (en) *  19930526  19950808  Kaplinsky; Cecil H.  Fast CMOS logic with programmable logic control 
US5349250A (en) *  19930902  19940920  Xilinx, Inc.  Logic structure and circuit for fast carry 
NonPatent Citations (2)
Title 

Z. Hasan, D. Harrison, and M. Ciesielski, "A Fast Partitioning Method for PLABased FPGA's" IEEE Design & Test of Computers, Dec. 1992, pp. 3439. 
Z. Hasan, D. Harrison, and M. Ciesielski, A Fast Partitioning Method for PLA Based FPGA s IEEE Design & Test of Computers, Dec. 1992, pp. 34 39. * 
Cited By (108)
Publication number  Priority date  Publication date  Assignee  Title 

US5963048A (en) *  19941223  19991005  Xilinx, Inc.  Method for programming complex PLD having more than one function block type 
US5926035A (en) *  19960626  19990720  Cypress Semiconductor Corp.  Method and apparatus to generate mask programmable device 
US5943488A (en) *  19960626  19990824  Cypress Semiconductor Corp.  Method and apparatus to generate mask programmable device 
US6118299A (en) *  19960626  20000912  Cypress Semiconductor Corp.  Method and apparatus to generate mask programmable device 
US6144220A (en) *  19960903  20001107  Xilinx, Inc.  FPGA Architecture using multiplexers that incorporate a logic gate 
US6091892A (en) *  19961113  20000718  Xilinx, Inc.  Method for mapping product terms in a complex programmable logic device 
US5790882A (en) *  19961113  19980804  Xilinx, Inc.  Programmable logic device placement method utilizing weighting function to facilitate pin locking 
US6195788B1 (en) *  19971017  20010227  Altera Corporation  Mapping heterogeneous logic elements in a programmable logic device 
US6453456B1 (en) *  20000322  20020917  Xilinx, Inc.  System and method for interactive implementation and testing of logic cores on a programmable logic device 
US6725441B1 (en)  20000322  20040420  Xilinx, Inc.  Method and apparatus for defining and modifying connections between logic cores implemented on programmable logic devices 
US7119576B1 (en)  20000918  20061010  Altera Corporation  Devices and methods with programmable logic and digital signal processing regions 
US7346644B1 (en)  20000918  20080318  Altera Corporation  Devices and methods with programmable logic and digital signal processing regions 
US7107382B2 (en) *  20030403  20060912  Emulex Design & Manufacturing Corporation  Virtual peripheral component interconnect multiplefunction device 
US20040199700A1 (en) *  20030403  20041007  Clayton Shawn Adam  Virtual peripheral component interconnect multiplefunction device 
US7350176B1 (en) *  20030717  20080325  Altera Corporation  Techniques for mapping to a shared lookup table mask 
US7249329B1 (en) *  20040601  20070724  Altera Corporation  Technology mapping techniques for incomplete lookup tables 
CN1329806C (en) *  20040818  20070801  å“ˆå°”æ»¨å·¥ä¸šå¤§å¦  Highspeed largecapacity data collecting system based on CPLD and SDRAM 
US20060247909A1 (en) *  20050427  20061102  Desai Madhav P  System and method for emulating a logic circuit design using programmable logic devices 
US8620980B1 (en)  20050927  20131231  Altera Corporation  Programmable device with specialized multiplier blocks 
US8041759B1 (en)  20060209  20111018  Altera Corporation  Specialized processing block for programmable logic device 
US8266199B2 (en)  20060209  20120911  Altera Corporation  Specialized processing block for programmable logic device 
US8266198B2 (en)  20060209  20120911  Altera Corporation  Specialized processing block for programmable logic device 
US8301681B1 (en)  20060209  20121030  Altera Corporation  Specialized processing block for programmable logic device 
US7836117B1 (en)  20060407  20101116  Altera Corporation  Specialized processing block for programmable logic device 
US7822799B1 (en)  20060626  20101026  Altera Corporation  Adderrounder circuitry for specialized processing block in programmable logic device 
US8386550B1 (en)  20060920  20130226  Altera Corporation  Method for configuring a finite impulse response filter in a programmable logic device 
US8788562B2 (en)  20061205  20140722  Altera Corporation  Large multiplier for programmable logic device 
US9395953B2 (en)  20061205  20160719  Altera Corporation  Large multiplier for programmable logic device 
US7930336B2 (en)  20061205  20110419  Altera Corporation  Large multiplier for programmable logic device 
US8386553B1 (en)  20061205  20130226  Altera Corporation  Large multiplier for programmable logic device 
US20080133627A1 (en) *  20061205  20080605  Altera Corporation  Large multiplier for programmable logic device 
US9063870B1 (en)  20061205  20150623  Altera Corporation  Large multiplier for programmable logic device 
US7814137B1 (en)  20070109  20101012  Altera Corporation  Combined interpolation and decimation filter for programmable logic device 
US7865541B1 (en)  20070122  20110104  Altera Corporation  Configuring floating point operations in a programmable logic device 
US8650231B1 (en)  20070122  20140211  Altera Corporation  Configuring floating point operations in a programmable device 
US8645450B1 (en)  20070302  20140204  Altera Corporation  Multiplieraccumulator circuitry and methods 
US7949699B1 (en)  20070830  20110524  Altera Corporation  Implementation of decimation filter in integrated circuit device using rambased data storage 
US8959137B1 (en)  20080220  20150217  Altera Corporation  Implementing large multipliers in a programmable integrated circuit device 
US8244789B1 (en)  20080314  20120814  Altera Corporation  Normalization of floating point operations in a programmable integrated circuit device 
US8886695B1 (en)  20080314  20141111  Altera Corporation  Normalization of floating point operations in a programmable integrated circuit device 
US8626815B1 (en)  20080714  20140107  Altera Corporation  Configuring a programmable integrated circuit device to perform matrix multiplication 
US8255448B1 (en)  20081002  20120828  Altera Corporation  Implementing division in a programmable integrated circuit device 
US8307023B1 (en)  20081010  20121106  Altera Corporation  DSP block for implementing large multiplier on a programmable integrated circuit device 
US8567946B2 (en)  20090213  20131029  Adlens Beacon, Inc.  Methods of filling a liquidfilled lens mechanism 
US8087778B2 (en)  20090213  20120103  Adlens Beacon, Inc.  Variable focus liquid filled lens mechanism 
US20110102738A1 (en) *  20090213  20110505  Adlens Beacon, Inc.  Variable Focus Liquid Filled Lens Apparatus 
US9033495B2 (en)  20090213  20150519  Adlens Beacon, Inc.  Variable focus liquid filled lens apparatus 
US8382280B2 (en)  20090213  20130226  Adlens Beacon, Inc.  Variable focus liquid filled lens apparatus 
US20100208194A1 (en) *  20090213  20100819  Amitava Gupta  Variable focus liquid filled lens apparatus 
US20100228806A1 (en) *  20090303  20100909  Keone Streicher  Modular digital signal processing circuitry with optionally usable, dedicated connections between modules of the circuitry 
US8706790B1 (en)  20090303  20140422  Altera Corporation  Implementing mixedprecision floatingpoint operations in a programmable integrated circuit device 
US8468192B1 (en)  20090303  20130618  Altera Corporation  Implementing multipliers in a programmable integrated circuit device 
US8645449B1 (en)  20090303  20140204  Altera Corporation  Combined floating point adder and subtractor 
US8751551B2 (en)  20090303  20140610  Altera Corporation  Modular digital signal processing circuitry with optionally usable, dedicated connections between modules of the circuitry 
US20100228807A1 (en) *  20090303  20100909  Altera Corporation  Digital signal processing circuitry with redundancy and bidirectional data paths 
US8620977B1 (en)  20090303  20131231  Altera Corporation  Modular digital signal processing circuitry with optionally usable, dedicated connections between modules of the circuitry 
US8886696B1 (en)  20090303  20141111  Altera Corporation  Digital signal processing circuitry with redundancy and ability to support larger multipliers 
US8549055B2 (en)  20090303  20131001  Altera Corporation  Modular digital signal processing circuitry with optionally usable, dedicated connections between modules of the circuitry 
US8805916B2 (en)  20090303  20140812  Altera Corporation  Digital signal processing circuitry with redundancy and bidirectional data paths 
US8650236B1 (en)  20090804  20140211  Altera Corporation  Highrate interpolation or decimation filter in integrated circuit device 
US8396914B1 (en)  20090911  20130312  Altera Corporation  Matrix decomposition in an integrated circuit device 
US8412756B1 (en)  20090911  20130402  Altera Corporation  Multioperand floating point operations in a programmable integrated circuit device 
US8414121B2 (en)  20091013  20130409  Adlens Beacon, Inc.  Nonround fluid filled lens optic 
US20110085243A1 (en) *  20091013  20110414  Amitava Gupta  NonRound Fluid Filled Lens Optic 
US8817381B2 (en)  20091013  20140826  Adlens Beacon, Inc.  Full field membrane design for nonround liquid lens assemblies 
US8136942B2 (en)  20091014  20120320  Adlens Beacon, Inc.  Aspheric fluid filled lens optic 
US20110085131A1 (en) *  20091014  20110414  Adlens Beacon, Inc.  Aspheric Fluid Filled Lens Optic 
US8760767B2 (en)  20091014  20140624  Adlens Beacon, Inc.  Fluid lens assembly 
US9354456B2 (en)  20091015  20160531  Adlens Beacon, Inc.  Hinge mechanism for a fluid filled lens assembly 
US8596781B2 (en)  20091015  20131203  Adlens Beacon, Inc.  Fluid filled lens reservoir system and manufacturing method of the reservoir system 
US20110102735A1 (en) *  20091015  20110505  Amitava Gupta  Fluid Filled Lens Reservoir System and Manufacturing Method of the Reservoir System 
US8876283B2 (en)  20091015  20141104  Adlens Beacon, Inc.  Hinge mechanism for a fluid filled lens assembly 
US8353593B2 (en)  20091015  20130115  Adlens Beacon, Inc.  Hinge mechanism for a fluid filled lens assembly 
US8708486B2 (en)  20091015  20140429  Adlens Beacon, Inc.  Fluid filled lenses and mechanisms of inflation thereof 
US8539016B1 (en)  20100209  20130917  Altera Corporation  QR decomposition in an integrated circuit device 
US7948267B1 (en)  20100209  20110524  Altera Corporation  Efficient rounding circuits and methods in configurable integrated circuit devices 
US8601044B2 (en)  20100302  20131203  Altera Corporation  Discrete Fourier Transform in an integrated circuit device 
US8458243B1 (en)  20100303  20130604  Altera Corporation  Digital signal processing circuit blocks with support for systolic finiteimpulseresponse digital filtering 
US8732225B1 (en)  20100303  20140520  Altera Corporation  Digital signal processing circuit blocks with support for systolic finiteimpulseresponse digital filtering 
US8589465B1 (en)  20100303  20131119  Altera Corporation  Digital signal processing circuit blocks with support for systolic finiteimpulseresponse digital filtering 
US8484265B1 (en)  20100304  20130709  Altera Corporation  Angular range reduction in an integrated circuit device 
US8510354B1 (en)  20100312  20130813  Altera Corporation  Calculation of trigonometric functions in an integrated circuit device 
US8539014B2 (en)  20100325  20130917  Altera Corporation  Solving linear matrices in an integrated circuit device 
US8812573B2 (en)  20100625  20140819  Altera Corporation  Calculation of trigonometric functions in an integrated circuit device 
US8862650B2 (en)  20100625  20141014  Altera Corporation  Calculation of trigonometric functions in an integrated circuit device 
US8589463B2 (en)  20100625  20131119  Altera Corporation  Calculation of trigonometric functions in an integrated circuit device 
US9036264B2 (en)  20100812  20150519  Adlens Beacon, Inc.  Fluidfilled lenses and their ophthalmic applications 
US8577951B1 (en)  20100819  20131105  Altera Corporation  Matrix operations in an integrated circuit device 
US8570658B2 (en)  20101011  20131029  Adlens Beacon, Inc.  Non powered concepts for a wire frame of fluid filled lenses 
US8488250B2 (en)  20101011  20130716  Adlens Beacon, Inc.  Perimeter piezo reservoir in a lens 
USD665009S1 (en)  20101014  20120807  Adlens Beacon, Inc.  Spectacles frame 
US9042027B2 (en)  20101110  20150526  Adlens Beacon, Inc.  Fluidfilled lenses and actuation systems thereof 
US8645451B2 (en)  20110310  20140204  Altera Corporation  Doubleclocked specialized processing block in an integrated circuit device 
US9600278B1 (en)  20110509  20170321  Altera Corporation  Programmable device using fixed and configurable logic to implement recursive trees 
US8812576B1 (en)  20110912  20140819  Altera Corporation  QR decomposition in an integrated circuit device 
US8949298B1 (en)  20110916  20150203  Altera Corporation  Computing floatingpoint polynomials in an integrated circuit device 
US9053045B1 (en)  20110916  20150609  Altera Corporation  Computing floatingpoint polynomials in an integrated circuit device 
US8762443B1 (en)  20111115  20140624  Altera Corporation  Matrix operations in an integrated circuit device 
US8543634B1 (en)  20120330  20130924  Altera Corporation  Specialized processing block for programmable integrated circuit device 
US9098332B1 (en)  20120601  20150804  Altera Corporation  Specialized processing block with fixed and floatingpoint structures 
US8996600B1 (en)  20120803  20150331  Altera Corporation  Specialized processing block for implementing floatingpoint multiplier with subnormal operation support 
US9207909B1 (en)  20121126  20151208  Altera Corporation  Polynomial calculations optimized for programmable integrated circuit device structures 
US9189200B1 (en)  20130314  20151117  Altera Corporation  Multipleprecision processing block in a programmable integrated circuit device 
US9348795B1 (en)  20130703  20160524  Altera Corporation  Programmable device using fixed and configurable logic to implement floatingpoint rounding 
US9846667B2 (en) *  20130809  20171219  Hon Hai Precision Industry Co., Ltd.  Switch apparatus and electronic device 
US20150046629A1 (en) *  20130809  20150212  Hon Hai Precision Industry Co., Ltd.  Switch apparatus and electronic device 
US9379687B1 (en)  20140114  20160628  Altera Corporation  Pipelined systolic finite impulse response filter 
US9684488B2 (en)  20150326  20170620  Altera Corporation  Combined adder and preadder for highradix multiplier circuit 
Also Published As
Publication number  Publication date  Type 

US5963048A (en)  19991005  grant 
Similar Documents
Publication  Publication Date  Title 

US6292019B1 (en)  Programmable logic device having configurable logic blocks with useraccessible input multiplexers  
US5359242A (en)  Programmable logic with carryin/carryout between logic blocks  
US6948147B1 (en)  Method and apparatus for configuring a programmable logic device using a master JTAG port  
US6633181B1 (en)  Multiscale programmable array  
US6359468B1 (en)  Programmable logic device with carry lookahead  
Hsieh et al.  Thirdgeneration architecture boosts speed and density of fieldprogrammable gate arrays  
US5581738A (en)  Method and apparatus for backannotating timing constraints into simulation models of field programmable gate arrays  
Maxfield  The design warrior's guide to FPGAs: devices, tools and flows  
US6943580B2 (en)  Fracturable lookup table and logic element  
US7075333B1 (en)  Programmable circuit optionally configurable as a lookup table or a wide multiplexer  
US5448493A (en)  Structure and method for manually controlling automatic configuration in an integrated circuit logic block array  
US7372297B1 (en)  Hybrid interconnect/logic circuits enabling efficient replication of a function in several subcycles to save logic and routing resources  
US6288570B1 (en)  Logic structure and circuit for fast carry  
US6998872B1 (en)  Lookup table circuit optionally configurable as two or more smaller lookup tables with independent inputs  
US20020010853A1 (en)  Method of time multiplexing a programmable logic device  
US6271680B1 (en)  Logic element for a programmable logic integrated circuit  
US5596743A (en)  Field programmable logic device with dynamic interconnections to a dynamic logic core  
US5835998A (en)  Logic cell for programmable logic devices  
US5399922A (en)  Macrocell comprised of two lookup tables and two flipflops  
Singh et al.  The effect of logic block architecture on FPGA performance  
US6421817B1 (en)  System and method of computation in a programmable logic device using virtual instructions  
US5815726A (en)  Coarsegrained lookup table architecture  
US7733123B1 (en)  Implementing conditional statements in selftimed logic circuits  
US6298472B1 (en)  Behavioral silicon construct architecture and mapping  
US6069490A (en)  Routing architecture using a direct connect routing mesh 
Legal Events
Date  Code  Title  Description 

AS  Assignment 
Owner name: XILINX, INC., CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:HARRISON, DAVID A.;SILVER, JOSHUA M.;SOE, SOREN T.;REEL/FRAME:007452/0678;SIGNING DATES FROM 19950224 TO 19950227 

AS  Assignment 
Owner name: XILINX, INC., CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:HARRISON, DAVID A.;SILVER, JOSHUA M.;SOE, SOREN T.;REEL/FRAME:008562/0651;SIGNING DATES FROM 19950224 TO 19950227 

FPAY  Fee payment 
Year of fee payment: 4 

FPAY  Fee payment 
Year of fee payment: 8 

FPAY  Fee payment 
Year of fee payment: 12 